Evaluation of Factors Affecting E-Bike Involved Crash and E-Bike License Plate Use in China Using a Bivariate Probit Model
Table 5
Estimate results of the BP model.
Variable
E-bike involved crash
E-bike license plate use
SE
value
95% conf. interval
SE
value
95% conf. interval
Constant
−2.132
0.877
0.015
−3.851
−0.413
−0.624
0.223
0.005
−1.061
−0.187
Gender (male versus female)
0.577
0.121
<0.001
0.340
0.814
−0.772
0.143
<0.001
−1.052
−0.492
Age group (young versus older)
0.362
0.135
0.007
0.097
0.627
−1.433
0.429
0.001
−2.274
−0.592
Education level (high versus low)
−0.530
0.254
0.037
−1.028
−0.031
1.542
0.577
0.008
0.411
2.673
Holding automobile driver license (yes versus no)
−0.619
0.125
<0.001
−0.864
−0.374
0.652
0.108
<0.001
0.440
0.864
Type of e-bike (E-scooter versus e-bike)
0.431
0.193
0.026
0.053
0.809
—
—
—
—
—
Car in household (yes versus no)
−0.445
0.142
0.002
−0.723
−0.167
0.371
0.121
0.002
0.134
0.608
Frequency of using e-bike (frequently versus occasionally)
0.511
0.216
0.018
0.088
0.934
—
—
—
—
—
Experiences in using e-bike (3–5 years versus <1 year)
−0.576
0.171
0.001
−0.911
−0.241
0.722
0.163
<0.001
0.403
1.041
Experiences in using e-bike (>5 years versus <1 year)
−0.466
0.142
0.001
−0.744
−0.188
0.574
0.171
0.001
0.239
0.909
Law compliance (strong versus no)
−1.115
0.376
0.003
−1.852
−0.378
1.221
0.196
<0.001
0.837
1.605
Aggressive driving behaviors (frequently versus no)
0.641
0.303
0.034
0.047
1.235
−0.232
0.106
0.029
−0.440
−0.024
Aggressive driving behaviors (generally versus no)
0.624
0.248
0.012
0.138
1.110
—
—
—
—
—
Impulse behavior (frequently versus no)
0.475
0.186
0.011
0.110
0.840
—
—
—
—
—
Degree of riding experience (skilled versus new driver)
−0.312
0.152
0.040
−0.610
−0.014
—
—
—
—
—
Risk perception scale (high versus low)
−0.287
0.114
0.012
−0.510
−0.064
—
—
—
—
—
Registration fee (low versus high)
—
—
—
—
—
0.511
0.221
0.002
0.078
0.944
Time for license registration (30 min versus >30 min)
—
—
—
—
—
0.654
0.236
0.001
0.191
1.117
Note. Number of observations = 862. The correlation of ( value < 0.001). Log likelihood at convergence = . — represents that the variable is not significant at a 95% confidence level.